AI Agent Governance: A Field Guide

TL;DR

This paper provides an accessible overview of AI agent governance, highlighting its importance, current state, and the need for societal preparedness for the widespread deployment of autonomous AI agents.

Contribution

It offers a comprehensive field guide to AI agent governance, emphasizing the urgency of developing policies and interventions for emerging autonomous AI systems.

Findings

AI agents could autonomously perform complex societal tasks

Society is largely unprepared for widespread autonomous agent deployment

Research on agent governance is still in early stages

Abstract

This report serves as an accessible guide to the emerging field of AI agent governance. Agents - AI systems that can autonomously achieve goals in the world, with little to no explicit human instruction about how to do so - are a major focus of leading tech companies, AI start-ups, and investors. If these development efforts are successful, some industry leaders claim we could soon see a world where millions or billions of agents autonomously perform complex tasks across society. Society is largely unprepared for this development. A future where capable agents are deployed en masse could see transformative benefits to society but also profound and novel risks. Currently, the exploration of agent governance questions and the development of associated interventions remain in their infancy. Only a few researchers, primarily in civil society organizations, public research institutes, and frontier AI companies, are actively working on these challenges.